Accelerating Digital Payments: Navi's Journey
summary
NAVI replaces a brittle sharded Postgres setup after index rebuilds trigger full table scans and 100% CPU, and after maintenance windows of up to 60 minutes. They adopt yugabyte managed for auto‑sharding, auto‑rebalancing and a smart cluster‑aware client, validate a sustained 2,000 TPS on a three‑node 8 vCPU cluster with zero‑downtime failovers, and report about 40% cost reduction versus vertical scaling; CDC and PG‑partman are pending managed features.
AI-written summary · about
chapters (10)
- 0:00Navi Technologies introduction
- 1:34Existing DB challenges
- 4:45New SQL database requirements
- 6:34Why YugabyteDB chosen
- 7:22Yugabyte performance and cost
- 8:45Developer experience with Yugabyte
- 9:50Connection scaling with Yugabyte
- 10:33Yugabyte security features
- 11:22Yugabyte future plans
- 12:26Closing remarks
related talks
20mPowering the Future of Payments with
18mTowards a Lightning Fast System: Migrating a Real-Time Payment Service to
11mChallenges Deploying Modern Real-Time Payment Systems on Traditional RDBMS
18mThe "New Normal" or An Evolution
10mBuilding High-Performance Financial Data Pipelines
29mDelivering Always-On Financial Services with
06mProject Lightning Talk: Vitess in Production: Powering Applications at Massive Sca...
30mStateful Superpowers: Explore High Performa...